Cheng Mingmei, Wang Jing, Liu Xiaofeng, Wang Yanzhong, Wu Qun, Wang Fangyun, Li Pei, Wang Binbin, Zhang Xin, Xie Wanqing
Department of Intelligent Medical Engineering, School of Biomedical Engineering, Department of Psychology, School of Mental Health and Psychological Sciences, Anhui Medical University, Hefei 230011, China.
Heart Center, Beijing Children's Hospital, Capital Medical University, National Center for Children's Health, Beijing 10045, China.
Research (Wash D C). 2024 Mar 6;7:0319. doi: 10.34133/research.0319. eCollection 2024.
Early detection and treatment of congenital heart disease (CHD) can significantly improve the prognosis of children. However, inexperienced sonographers often face difficulties in recognizing CHD through transthoracic echocardiogram (TTE) images. In this study, 2-dimensional (2D) and Doppler TTEs of children collected from 2 clinical groups from Beijing Children's Hospital between 2018 and 2022 were analyzed, including views of apical 4 chamber, subxiphoid long-axis view of 2 atria, parasternal long-axis view of the left ventricle, parasternal short-axis view of aorta, and suprasternal long-axis view. A deep learning (DL) framework was developed to identify cardiac views, integrate information from various views and modalities, visualize the high-risk region, and predict the probability of the subject being normal or having an atrial septal defect (ASD) or a ventricular septaldefect (VSD). A total of 1,932 children (1,255 healthy controls, 292 ASDs, and 385 VSDs) were collected from 2 clinical groups. For view classification, the DL model reached a mean [SD] accuracy of 0.989 [0.001]. For CHD screening, the model using both 2D and Doppler TTEs with 5 views achieved a mean [SD] area under the receiver operating characteristic curve (AUC) of 0.996 [0.000] and an accuracy of 0.994 [0.002] for within-center evaluation while reaching a mean [SD] AUC of 0.990 [0.003] and an accuracy of 0.993 [0.001] for cross-center test set. For the classification of healthy, ASD, and VSD, the model reached the mean [SD] accuracy of 0.991 [0.002] and 0.986 [0.001] for within- and cross-center evaluation, respectively. The DL models aggregating TTEs with more modalities and scanning views attained superior performance to approximate that of experienced sonographers. The incorporation of multiple views and modalities of TTEs in the model enables accurate identification of children with CHD in a noninvasive manner, suggesting the potential to enhance CHD detection performance and simplify the screening process.
先天性心脏病（CHD）的早期检测和治疗可显著改善儿童的预后。然而，经验不足的超声心动图检查人员在通过经胸超声心动图（TTE）图像识别CHD时常常面临困难。在本研究中，对2018年至2022年期间从北京儿童医院的2个临床组收集的儿童二维（2D）和多普勒TTE进行了分析，包括心尖四腔心切面、两心房剑突下长轴切面、左心室胸骨旁长轴切面、主动脉胸骨旁短轴切面和胸骨上窝长轴切面。开发了一种深度学习（DL）框架，用于识别心脏切面、整合来自各种切面和模式的信息、可视化高危区域，并预测受试者正常、患有房间隔缺损（ASD）或室间隔缺损（VSD）的概率。从2个临床组共收集了1932名儿童（1255名健康对照、292名ASD患儿和385名VSD患儿）。对于切面分类，DL模型的平均[标准差]准确率达到0.989[0.001]。对于CHD筛查，使用2D和多普勒TTE的5个切面的模型在中心内评估时，受试者操作特征曲线（AUC）下的平均[标准差]面积为0.996[0.000]，准确率为0.994[0.002]；而在跨中心测试集中，平均[标准差]AUC为0.990[0.003]，准确率为0.993[0.001]。对于健康、ASD和VSD的分类，该模型在中心内和跨中心评估中的平均[标准差]准确率分别达到0.991[0.002]和0.986[0.001]。聚合更多模式和扫描切面的TTE的DL模型表现优于经验丰富的超声心动图检查人员。该模型中纳入TTE的多个切面和模式能够以非侵入性方式准确识别患有CHD的儿童，表明具有提高CHD检测性能和简化筛查过程的潜力。
